Approach to turn off fan of Z11 / Z12 / Z13

Hello everybody.

My aim is to turn off the fan of my Z12 in low load situations like when reading / writing documents, because the rather quiet but still noticable noise of the fan bothers me if I use the Z12 in a silent environment. For this I want to write a program which monitors the temperature and load and turns off the fan if it is possible.

This is what I know:
Will this result in harmful temperatures / stability issues? No.
Is it possible to turn off the fan without modifying the hardware? Yes.

Temperature tests:
I have run my Z12 with the fan turned off for 10 minutes with minimal CPU load and all unneeded devices turned off during the test. No stability issues occured. Here are the monitored temperatures (in °C) of the sensors. For comparison a 1-minute-prime95-run (with fan at max speed) after cooling the system down again.

Code:

Sensor    idle (fan low)   idle (fan off, 10 min.)  prime 95 (fan max, 1 min.)
TP50      51               62                       83
CPU       51/49            63/61                    83/82
HDD       31               32                       33

As expected all tempertures have increased during the test with the fan turned off. But they all are in a feasible region after the 10 minutes and far below the temperatures of the prime95-run with the fan turned on. This shows that a completely passive cooling in a rather cool environment (approx 20°C) is possible.

How to turn off the fan:
I have discovered, that it is possible to turn off the fan of my Z12 by switching the temperature control modes quickly. After a few seconds the fan turns off completely. Sadly, I have to keep on pressing the button every second to prevent it from turning on again. That's exactly what I have done during the 10-minute temperature test mentioned before. That's why I am looking for a software approach to achieve this.

My question:
Does anyone have an idea how to toggle the temperature control modes by software in order to turn off the fan?


========================================= Here the original post:

Hi folks,

I just discovered, that it is possible to turn off the fan of my Z12 by switching the temperature control modes quickly for a while. After a few seconds the fan turns off completely. Sadly, I have to keep on pressing the button every second to prevent it from turning on again. Therefore I am looking for a software based approach.

To be percise, I am interested in creating a tool which turns off the fan as long as the system temperature is very low. Sadly, I could not find out how to change the temperature control modes via software. But this seems to be a promising approach.

Any ideas ... ?

 

Why bother? It's silent at low speed. If yours isn't, something's wrong with it.

 

I think you can do that with the round button that's built into the right hand side of the hinge.
Let me know if it works for you.

 

It's rather silent in energy saving / silent mode but it still bothers me if working in a quiet environment.

I would like it to be cooled passively if writing documents. This should work out in a cool environment since the power demand is very low if the cpu is throttled, the intel graphics is used, the battery is loaded, and an external monitor is used as it is in my case.

 

Yes, I have mapped the change-thermal-control-function to the s-button. But the fan turns off completely only if I press this button every (!) second. I could design an electric circuit to achieve this but a software solution is preferable, of course.

The first step is to find out how to achieve this (e.g. by registry setting / launching a sony application / ...).
Then I will try to supress the notification message and possible disk accesses which occur if the thermal-control-mode is changed manually.

 

Really you shouldn't hear it at all under low load.

 

sorry olli - my attempt at sarcasm was missed - I was suggesting that you use te power button - it works every time in shutting up that pesky fan.

I can understand what you want but I think it's a bit silly. The thing creates heat as part of it's core function, just like an incandescent globe. There are a few different ways of dissipating that heat and typically with a laptop it's through the use of moving air to get the heat away from (relatively) fragile components). Generally the more powerful the laptop is, the more air you have to move.
If you're that keen on silent operation then I think you've got the wrong computer.
Have fun with you fiddling.

 

The effectiveness of the cooling, and consequently the speed of the fan, depends on the temperature of the air going in. Reduce the ambient temperature and the fan sounds should reduce all by themselves. You'll catch pneumonia, but that's a mere side effect.

For desktops we are able to do weird stuff like water cooling - kind of hard to do that with portable devices.

 

the idea is absolutely useless. If you are willing to experiment, disconnect the fan and you will notice how fast the temp. of the laptop rises! So basically there is no "as long as the system temperature is very low", as this will be a period of about 2-3 seconds. The constant on- and off-going of a fan is much more disturbing, than having it spin on its lowest speed!

It not only spins to cool down - it also keeps a nearly constant temperature environment this way. If you do stop it - the temperature will very quickly rise!

 

I am using a completely passively cooled mini-server since 5 years without problems. Its heat dissipation is about 10 W and its body is a lot smaller than the Z12.

In the scenario I mentioned, I guess the minimum power demand of the Z12 in active state is about 10-15 W and the heat dissipating aluminium body is rather large. So it should work out. Of course, as soon as the power demand increases the fan must spin up - but that's ok then.

90% of the time I'm using the Z12 as desktop replacement only reading / writing documents in a silent environment. That's why I am interested in turning the fan completely off.

 

I don't know your mini-server, but again - just try to disconnect or block (don't know if this would work) your fan for a minute and observe how the temperature will increase drastically even if you don't do anything!

Or simulate by putting the laptop in a really cool environment so that the FAN would stop spinning by itself once the temp goes low. Then you will again notice how the temp increases rapidly without a fan!

What you dream of can never work on the Z.

 

I think you're all being a bit harsh.

Firstly, the fan on Vaio Z is not silent. Of course, it depends on your environment, the ambient noise level, and how good your hearing is. I have had three different Vaio Z and they are all the same - in a quiet environment, the fan is very noticeable, with an audible whine. There is a definite pitch to the whine as if you put two Vaio Zs next to each other, you can hear the two fans going in and out of sync.

Secondly, a number of portables work without a fan running for a period of time (more than a few seconds), without them melting into oblivion. Though I do appreciate the vaio Z packs a lot into a small space which does mean temperatures can rise more quickly.

Thirdly, tools do exist on other PCs that allow you adjust when the fan runs and how fast it runs, such as TPFanControl for the thinkpad.

So I don't think it was an unreasonable request. However, in answer to the original question, I don't know of any tools that do that. And I agree, it would be nice if the PC was a little quieter!

 

Why can't we thermally link the aluminium keypad bezel to the internal heatsinks? Then the entire top surface will dissipate heat. IMO that will be a much more efficient way of keeping temperatures down when the fan is off or running slowly as desired.

 

... and would keep a cup of coffee warm, when put on top. Very practical and not at all unpleasant for the user ......

 

Sarcasm uncalled for. Firstly, the temperature won't go as high. Secondly, who touches the aluminium top? The palm rest is some kind of plastic.

 

Always when I have such revolutionary ideas I ask myself "Do you reaaaaaalllly think no one thought about that before and there is no reason for something being the way it is?"

But if you have free time and don't care about the laptop itself - try that I'd love to see the results (No sarcasm here! )

 

Who touches the aluminum top??? Anyone who types?

Thermally coupling the c panel could be troublesome for a number of reasons. Big one being there are many plastic parts in the keyboard assemble (such as the keys themselves) that likely cannot manage that type of heat. Even if they can, the keys would get quite hot when the cpu and gpu are under load.

Maybe I just got a lucky Z, but mine is silent. Silent in a silent environment. If I put my ear up to the exhaust port all I hear is moving air, and yes I have good hearing.

I'm no programmer, but if repeatedly changing the modes accomplishes what o/p wants, it seems you could just write a script to keep executing whatever program is being called when you push the button that you can map to the thermal strategy. Seems easy enough. Then start enjoying the thermal shutdowns.

 

The keys are not in contact with the top. The top has cutouts.

After your Z has run for a while on load, touch the top on the left hand side (just above the exhaust vent). Is it warm? Is the warmth causing major problems? Now touch the top on the right hand side. It is cool - not dissipating the heat it might have, had the designers thought about it.

P.S. You must have huge fingers - I can type without actually touching the metal surface, just the keytops.

 

That script is exactly what I want to accomplish. The problem is, that I don't know "whatever program is being called".

This script should, of course, monitor the temperatures so that thermal shutdowns should not occur.

 

You are right. When typing you usually touching only plastic parts (the keys and the palmrest). Due to the low thermal conductivity of these parts the surface temperture is not critical and will quickly adopt the users skin temperature when being touched. I like the idea of using more of the aluminium body for heat dissipation.

 

I think at least the GPU VRAM seems to dissipate heat there (rather, to the bottom of the keyboard):
ImageShack® - Online Photo and Video Hosting
http://forum.notebookreview.com/6009997-post86.html

 

I hope you don't really mean the above!?

so the top is just "hanging" in the air? Of course it is in contact with many plastic parts, even if not the keys themselves, at least not directly, but they will get warm as well.
And what do you plan to do when loading the CPU&GPU - "disconnect" the top manually or let it get as hot as it does and melt plastics?

I'm not saying the idea is stupid - more or less this is what Apple are trying to accomplish as well. But it needs much more engineering than you can imagine Besides that, heat transportation isn't as easy as electricity conduction - not every metal is a good heat conductor, not to mention that paint doesn't like heat. Notice that heatsinks are made of specific metal alloys.
The point is pretty much every metal dissipates heat to some extend. But not every metal has the ability to conduct it further and thus use the entire available area. Furthermore, your connection needs to have the corresponding qualities and size to be able to conduct any heat at all. You don't expect to help the cooling by connecting your heatsink to the top with a single wire, do you?

Heat is more complicated than electricity

 

Perhaps I should have mentioned that I am an electronics hardware design engineer with over 25 years experience. Aluminium is a very common material for heatsinks, copper being considered too heavy for many applications. Look at any large audio amplifier (as a common example) and note the black finned heatsinks at the back - those are aluminium, anodised black for looks and corrosion resistance.

The chips start throttling around 90 deg C. Thermal resistance from the chips to the top plate (you can look that up if you like) will ensure that the top will not go above 60 deg C. Most plastics will handle that without fuss, even for long durations.

 

Actually the electric conduction and the conduction of heat are physically absolutely analogous. That's why e. g. identical equation systems are used when simulating such problems.

For the problem at hand, I think following is of interest:
Firstly, the heat conductivity of the aluminium is far, far higher than that of the plasic parts. Which is perfect, since we would like to dissipate the heat through the untouched aluminium body.
Secondly, the maximum tolerabel temperature of the plastic parts should be far higher than anything that would appear in practice even if directly connected to the heat sinks. Typically melting temperatures of thermoplastic parts are above 200°C. The termally most sensitive parts in computers are usually the capacitors and maybe some glue or rubber parts.

 

Ok guys. Make it work then I will be following the thread with great interest to see the result.

I have no 25 years experience, but my quite extend knowledge in thermodynamics tells me it's a lost cause unless you redesign the whole thermal concept and basically create a new body.

If it was that easy, it would have been made already. And it actually has - both by Sony and Apple. The body is used for dissipation but it cannot replace a running fan, especially for models with dedicated GPU which dissipates more heat than the CPU.

 

No one said it would replace a running fan with everything running full blast. The OP wanted something ultra quiet while he was doing very light stuff, like reading an ebook. IMO this approach fits the bill, but whether or not it can be implemented will depend almost totally on what we find when we open up the case. If there is no easy way to improve the thermal path between the main internal heatsinks and heat pipes and the aluminium top, then this approach is dead in the water, and will have to be left to laptop design teams.

Of course, when the system load increases, the fans will have to come on to get rid of the excess heat. I'm not advocating totally passive cooling under all conditions, just in the special case when the system load is minimal.

 

^I think if sony felt the machine could maintain safe temps under even ZERO load, they would've programmed the fan to spin down to zero when available. And no matter how you phrase it, thermally coupling to the C panel is going to give the undesirable side effect of making the keyboard and surrounding area warm (or hot).

If your fan truly makes noise at it's lowest setting, then I can imagine the constant spinning up of the fan from off to HIGH will drive you just as crazy, assuming you get this to work in the first place.

To test the theory, why not just disconnect the fan and lightly use the laptop? I bet it overheats within a few minutes.

 

That's exactly what I said. And even if we imagine using a part of the laptop for better heat dissipation - this will not prevent it from overheating. It will just delay it with a second or so.

Achieving a condition where the FAN goes on and off every 30 seconds will bother you MUCH more, than having it spin on its lowest speed all the time I know this from experience

 

Use it in the snowy desert... best solution..no fan required!

but seriously. ..guys I think you're obsessing over laptop fans. get over it. There is no technical solution today to run microprocessors in a thin closed box without a fan at room temperatures.

 

OK, that's what I have done. I have performed a temperature and stability test with the fan turned off for 10 minutes. There were no thermal and stability issues at all:

Passively cooling a Z12 at low load is definitely possible!

I have added the test results to the initial post.

 

In regards to the OPs original post...

I have a VGN-Z and (somewhat) found out about the fan's triggers during some troubleshooting a while back. You may be able to map the BIOS ROM into a VPC-Z BIOS update file or something; what you are asking will require you to essentially 'reprogramme' the mobo. Though unsure if that's a wise idea considering the internals of a VGN-Z and VPC-Z are drastically different. Well, that plus, I don't know if mapping parts of a BIOS into another model's BIOS is even possible, let alone safe if you are successful...

So the fan on the VGN-Z:
- Turns COMPLETELY off (ie, stops) when temps hit and go lower than 30'C
- Turns ON into SLOW mode when temps hit 37'C
- Enters the SLOW-MID speeds when it hits 45'C
- Enters the MID speeds when it hits 60'C
- Enters the MID-HI speed when hits 75'C
- Goes into FULL speed when >80'C (sometimes delayed till when >85'C)

Notes:
- The same but backwards when the temps start decreasing.
- The above applies for both CPU and GPU; if any one of those components hits the above temps, the fan will spin at said fan speeds. And when decreasing, fan slows to said speed when BOTH components meet the prerequistes.

I found this out researching the fan-patterns a year or so back, as my Z makes a funky high pitched sound when it attempts to spin the fan from a stopped state to a moving state (99% of the time; sound seldom doesn't occur). Though this sound in question doesn't occur when the fan is changing speeds when already moving. I took my Z apart to see what the fan does, and whever it makes the sound the fan looks like it's stalling itself into moving (if that makes sense). Though the fan always manages to start; never had the fan NOT start due to the sound and / or whatever is causing it.

That said, I too mapped an S button to toggle between SILENT, BALANCE and HIGH PERFORMANCE. If I toggle to HIGH while the fan is stopped, the sound doesn't occur and blasts into MID-HIGH speeds without a complaint. So yeah... Unsure what's going on with my machine. No complaints though as it's been like this for a year and a half and still going (not a single overheat since purchase).

Good luck with your wee project.

 

Thank you for this interesting insight into the VGN-Z. VGN-Z and VPC-Z are rather comparible in power demand and size. Sony did support turning off the fan for the VGN-Z as you mentioned. Maybe the issues you described when turning the fan on again are the reason why the VPC-Z doesn't support this anymore.

The idea of patching the BIOS sounds rather challenging. I am not able to achieve this. But I hope that a less invasive software solution is possible. Actually all I need is someone who can tell me how to toggle the modes by software. The sony software does so, therefore it must be possible.

 

^Wasn't VGN-Z 25W cpu?

 

For the comparison of thermal properties of two systems the TDP of the CPU is only of little use.

On one side the power demand of mainboard / GPU / voltage converters / ... are missing. On the other side the TDP is much higher than the power demand of the CPU if it is idle.

Measuring the overall power demand of the laptop plus external power converter is all I can do without special equipment. The batteries should, of course, be either unplugged or completely charged.